Abstract:Benefit from the promising features of second-order correlation, ghost imaging (GI) has received extensive attentions in recent years. Simultaneously, GI is affected by the poor trade-off between sampling rate and imaging quality. The traditional image reconstruction method in GI is to accumulate the action result of each speckle and the corresponding bucket signal. We found that the image reconstruction process of GI is very similar to the Recurrent Neural Network (RNN), which is one of the deep learning algorithm. In this paper, we proposed a novel method that effectively implements GI on the RNN architecture, called GI-RNN. The state of each layer in RNN is determined by the output of the previous layer and the input of this layer, and the output of the network is the sum of all previous states. Therefore, we take the speckle of each illumination and the corresponding bucket signal as the input of each layer, and the output of the network is the sum of all previous speckle and bucket signal, which is the image of the target. The testing results show that the proposed method can achieve image reconstruction at a very low sampling rate (0.38$\%$). Moreover, we compare GI-RNN with traditional GI algorithm and compressed sensing algorithm. The results of different targets show that GI-RNN is 6.61 dB higher than compressed sensing algorithm and 12.58 dB higher than traditional GI algorithm on average. In our view, the proposed method makes an important step to applications of GI.

Multidirectional Edge Detection Based on Gradient Ghost Imaging

Corner Detection Based on Gradient Ghost Imaging

Compressive sensing ghost imaging based on image gradient

Edge detection in gradient ghost imaging through complex media

Ghost Imaging with N-order Speckle Light

Adaptive Polarization Compressive Ghost Imaging Based on Principal Component Analysis

Compressive Sensing Ghost Imaging Object Detection Using Generative Adversarial Networks

Foveated Ghost Imaging Based on Deep Learning

Compressive Sensing Ghost Imaging Based On Neighbor Similarity

Ghost Difference Imaging Using One Single-Pixel Detector

Omnidirectional ghost imaging system and unwrapping-free panoramic ghost imaging

Time-correspondence differential ghost imaging

Dual-mode adaptive-SVD ghost imaging

Compressed ghost edge imaging

Joint Demosaicing and Denoising with Gradient Guidance in Quad Bayer CFA

Multispectral computational ghost imaging with multiplexed illumination

Ultrasound Image Edge Detection Based on a Novel Multiplicative Gradient and Canny Operator.

Edge detection based on joint iteration ghost imaging

Instant ghost imaging: improving robustness for ghost imaging subject to optical background noise

Ghost Imaging Based on Recurrent Neural Network

Double filter iterative ghost imaging for high quality edge and image acquisition